Recent studies determined that antidepressives of new generation were inadequate for moderate or severe depressions. The improvement these diseases show with medication or with a placebo isn't very different, according to statistics. Antidepressives seem to be efficient only upon patients that present an extreme depression. Nevertheless, the apparent efficiency upon this group wouldn't be due to the medication itself, but to a minor response to the placebo. Another group of researchers came to very different conclusions employing a similar methodology. They assured that all antidepressives were superior compared to the placebo, although their real effect would be even minor if analysis of studies not yet published were included. This situation shows the difficulties the investigation on psychoremedies for depressions present and it revives the polemics about the utility and the ethical justification of employing placebos in this type of studies.